On Fri, Jan 09, 2004 at 02:36:33AM +0000, MJ Ray wrote: > For me, at least, I think nonfree.org is a forseeable consequence of > the GR, but I still would not think it ethical to support it because I > believe it does not help our users, long-term. It's just an answer to > the question "what will happen to current non-free?" that seems to > worry some people.
If it's unethical and harmful to our users, then it's _not_ an answer to
the question that worries people: "will this benefit or harm our users?"
If nonfree.org *will* happen, then the hypothetical benefits to our users
due to more work happening on free software because of the lack of non-free
software aren't going to come to pass.
And by contrast, if you want nonfree.org *not* to happen -- if you
want to actually get the benefits that you seem to be envisaging -- you
need to find some other way of addressing the concerns that nonfree.org
would solve.
The arguments you and others are making here just aren't consistent.
> Meanwhile, I don't understand how people can argue that there is
> nothing special about non-free being on the debian systems, yet a much
> smaller-scale version of a similar infrastructure is beyond the rest
> of the world.
The infrastructure work involved in maintaining Debian as it is is:
M: Support "main" component for 10k packages
X: Support multiple components for an additional 1k packages
The infrastructure work involved in maintaining non-free.org is:
N: Support "non-free" component for 1k packages
"M" is large, "X" is trivial by comparison. The difference between "M" and
"N" doesn't seem particularly great, and "N" seems consequently much larger
than "X".
Currently, we support non-free at a cost of M+X.
In the future, we either support contrib at a cost of M+X, or we just support
main at a cost of M.
If nonfree.org gets created and supported, it's maintained at a cost of N.
So the outcomes are:
* Keep contrib, setup nonfree.org: cost M+X+N
* Keep contrib, nonfree.org not needed: cost M+X
* Drop contrib, setup nonfree.org: cost M+N
* Drop contrib, nonfree.org not needed: cost M
The only one of these that reduces our time spent on maintaining
infrastructure is the latter; all the others maintain it, or increase
it substantially.
I don't think anyone's arguing that it's beyond the rest of the world;
personally I'm just arguing that it's a waste of effort for no gain;
or at the very least that the costs are significant and definite, and
the gains are insubstantial and hypothetical.
It's entirely possible to prove me wrong here, if that's actually the
case: do the experiment and measure the time and effort it takes. The
effort that'd go into that is around "N" too -- so if you think it's
too much effort, well, doesn't that just mean you agree with me?
Cheers,
aj
--
Anthony Towns <[EMAIL PROTECTED]> <http://azure.humbug.org.au/~aj/>
I don't speak for anyone save myself. GPG signed mail preferred.
Linux.conf.au 2004 -- Because we can.
http://conf.linux.org.au/ -- Jan 12-17, 2004
signature.asc
Description: Digital signature
